Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
parquet
Sub-tasks:
acceptability-classification
Languages:
English
Size:
10K - 100K
ArXiv:
License:
Commit
•
68f498c
1
Parent(s):
e7a8756
Add tough_vs_raising_2 data files
Browse files- README.md +7 -3
- dataset_infos.json +6 -24
- tough_vs_raising_2/train-00000-of-00001.parquet +3 -0
README.md
CHANGED
@@ -1640,10 +1640,10 @@ dataset_info:
|
|
1640 |
dtype: int32
|
1641 |
splits:
|
1642 |
- name: train
|
1643 |
-
num_bytes:
|
1644 |
num_examples: 1000
|
1645 |
-
download_size:
|
1646 |
-
dataset_size:
|
1647 |
- config_name: transitive
|
1648 |
features:
|
1649 |
- name: sentence_good
|
@@ -2125,6 +2125,10 @@ configs:
|
|
2125 |
data_files:
|
2126 |
- split: train
|
2127 |
path: tough_vs_raising_1/train-*
|
|
|
|
|
|
|
|
|
2128 |
---
|
2129 |
|
2130 |
# Dataset Card for "blimp"
|
|
|
1640 |
dtype: int32
|
1641 |
splits:
|
1642 |
- name: train
|
1643 |
+
num_bytes: 169684
|
1644 |
num_examples: 1000
|
1645 |
+
download_size: 61465
|
1646 |
+
dataset_size: 169684
|
1647 |
- config_name: transitive
|
1648 |
features:
|
1649 |
- name: sentence_good
|
|
|
2125 |
data_files:
|
2126 |
- split: train
|
2127 |
path: tough_vs_raising_1/train-*
|
2128 |
+
- config_name: tough_vs_raising_2
|
2129 |
+
data_files:
|
2130 |
+
- split: train
|
2131 |
+
path: tough_vs_raising_2/train-*
|
2132 |
---
|
2133 |
|
2134 |
# Dataset Card for "blimp"
|
dataset_infos.json
CHANGED
@@ -3883,62 +3883,50 @@
|
|
3883 |
"features": {
|
3884 |
"sentence_good": {
|
3885 |
"dtype": "string",
|
3886 |
-
"id": null,
|
3887 |
"_type": "Value"
|
3888 |
},
|
3889 |
"sentence_bad": {
|
3890 |
"dtype": "string",
|
3891 |
-
"id": null,
|
3892 |
"_type": "Value"
|
3893 |
},
|
3894 |
"field": {
|
3895 |
"dtype": "string",
|
3896 |
-
"id": null,
|
3897 |
"_type": "Value"
|
3898 |
},
|
3899 |
"linguistics_term": {
|
3900 |
"dtype": "string",
|
3901 |
-
"id": null,
|
3902 |
"_type": "Value"
|
3903 |
},
|
3904 |
"UID": {
|
3905 |
"dtype": "string",
|
3906 |
-
"id": null,
|
3907 |
"_type": "Value"
|
3908 |
},
|
3909 |
"simple_LM_method": {
|
3910 |
"dtype": "bool",
|
3911 |
-
"id": null,
|
3912 |
"_type": "Value"
|
3913 |
},
|
3914 |
"one_prefix_method": {
|
3915 |
"dtype": "bool",
|
3916 |
-
"id": null,
|
3917 |
"_type": "Value"
|
3918 |
},
|
3919 |
"two_prefix_method": {
|
3920 |
"dtype": "bool",
|
3921 |
-
"id": null,
|
3922 |
"_type": "Value"
|
3923 |
},
|
3924 |
"lexically_identical": {
|
3925 |
"dtype": "bool",
|
3926 |
-
"id": null,
|
3927 |
"_type": "Value"
|
3928 |
},
|
3929 |
"pair_id": {
|
3930 |
"dtype": "int32",
|
3931 |
-
"id": null,
|
3932 |
"_type": "Value"
|
3933 |
}
|
3934 |
},
|
3935 |
-
"supervised_keys": null,
|
3936 |
"builder_name": "blimp",
|
|
|
3937 |
"config_name": "tough_vs_raising_2",
|
3938 |
"version": {
|
3939 |
"version_str": "0.1.0",
|
3940 |
-
"description": null,
|
3941 |
-
"datasets_version_to_prepare": null,
|
3942 |
"major": 0,
|
3943 |
"minor": 1,
|
3944 |
"patch": 0
|
@@ -3946,20 +3934,14 @@
|
|
3946 |
"splits": {
|
3947 |
"train": {
|
3948 |
"name": "train",
|
3949 |
-
"num_bytes":
|
3950 |
"num_examples": 1000,
|
3951 |
-
"dataset_name":
|
3952 |
-
}
|
3953 |
-
},
|
3954 |
-
"download_checksums": {
|
3955 |
-
"https://raw.githubusercontent.com/alexwarstadt/blimp/master/data/tough_vs_raising_2.jsonl": {
|
3956 |
-
"num_bytes": 364074,
|
3957 |
-
"checksum": "a6db2f24a8aee0c90a9be63e6bb6787fe04e990c518761115f69f27c7438b5ae"
|
3958 |
}
|
3959 |
},
|
3960 |
-
"download_size":
|
3961 |
-
"dataset_size":
|
3962 |
-
"size_in_bytes":
|
3963 |
},
|
3964 |
"transitive": {
|
3965 |
"description": "\nBLiMP is a challenge set for evaluating what language models (LMs) know about\nmajor grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each\ncontaining 1000 minimal pairs isolating specific contrasts in syntax,\nmorphology, or semantics. The data is automatically generated according to\nexpert-crafted grammars.\n",
|
|
|
3883 |
"features": {
|
3884 |
"sentence_good": {
|
3885 |
"dtype": "string",
|
|
|
3886 |
"_type": "Value"
|
3887 |
},
|
3888 |
"sentence_bad": {
|
3889 |
"dtype": "string",
|
|
|
3890 |
"_type": "Value"
|
3891 |
},
|
3892 |
"field": {
|
3893 |
"dtype": "string",
|
|
|
3894 |
"_type": "Value"
|
3895 |
},
|
3896 |
"linguistics_term": {
|
3897 |
"dtype": "string",
|
|
|
3898 |
"_type": "Value"
|
3899 |
},
|
3900 |
"UID": {
|
3901 |
"dtype": "string",
|
|
|
3902 |
"_type": "Value"
|
3903 |
},
|
3904 |
"simple_LM_method": {
|
3905 |
"dtype": "bool",
|
|
|
3906 |
"_type": "Value"
|
3907 |
},
|
3908 |
"one_prefix_method": {
|
3909 |
"dtype": "bool",
|
|
|
3910 |
"_type": "Value"
|
3911 |
},
|
3912 |
"two_prefix_method": {
|
3913 |
"dtype": "bool",
|
|
|
3914 |
"_type": "Value"
|
3915 |
},
|
3916 |
"lexically_identical": {
|
3917 |
"dtype": "bool",
|
|
|
3918 |
"_type": "Value"
|
3919 |
},
|
3920 |
"pair_id": {
|
3921 |
"dtype": "int32",
|
|
|
3922 |
"_type": "Value"
|
3923 |
}
|
3924 |
},
|
|
|
3925 |
"builder_name": "blimp",
|
3926 |
+
"dataset_name": "blimp",
|
3927 |
"config_name": "tough_vs_raising_2",
|
3928 |
"version": {
|
3929 |
"version_str": "0.1.0",
|
|
|
|
|
3930 |
"major": 0,
|
3931 |
"minor": 1,
|
3932 |
"patch": 0
|
|
|
3934 |
"splits": {
|
3935 |
"train": {
|
3936 |
"name": "train",
|
3937 |
+
"num_bytes": 169684,
|
3938 |
"num_examples": 1000,
|
3939 |
+
"dataset_name": null
|
|
|
|
|
|
|
|
|
|
|
|
|
3940 |
}
|
3941 |
},
|
3942 |
+
"download_size": 61465,
|
3943 |
+
"dataset_size": 169684,
|
3944 |
+
"size_in_bytes": 231149
|
3945 |
},
|
3946 |
"transitive": {
|
3947 |
"description": "\nBLiMP is a challenge set for evaluating what language models (LMs) know about\nmajor grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each\ncontaining 1000 minimal pairs isolating specific contrasts in syntax,\nmorphology, or semantics. The data is automatically generated according to\nexpert-crafted grammars.\n",
|
tough_vs_raising_2/train-00000-of-00001.parquet
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:01d9c6a6b0df268db59ce28ef6904541df2c0755cb5915b2f1f5dbcf08c3d82a
|
3 |
+
size 61465
|